Process Monitoring and Steady State Identification for Start-up Process of Batch Processes

Abstract

Inter-batch evolution widely exists in the start-up process of batch processes, and different operation phases exhibit significantly different behaviors, so multiple models are necessary both along the batch direction and the time direction. In this work, reference windows and sliding windows are constructed within each phase for analysis of inter-batch evolution. Through analyzing two indexes, Ratio and , the process is divided into different operating states. Different models are constructed for these states. In addition, a PCA similarity factor and the Mahalanobis distances between batch trajectories are used for online steady-state identification (SSID). The application to a typical multiphase batch process with inter-batch evolution, injection molding start-up process, illustrates the feasibility and performance of the proposed algorithm.